INTRODUCTION

Top Abstract Introduction Materials and methods Results and discussion References

Noncomplementary basepairing in DNA occurs either due to replication error, during recombination between homologous but nonidentical DNA sequences, or due to chemical modification of bases such as deamination of 5-methylcytosine to thymine. These mismatches, if not repaired, result in a high spontaneous mutation frequency. Several mechanisms have been proposed for the repair of mismatches in DNA (11, 20, 21). The methyl-directed postreplicational mismatch repair pathway involves the functions of the dam, mutS, mutL, mutH, and mutU gene products. This repair pathway recognizes the state of methylation of each DNA strand at d(GATC) sequences and acts on the strand which is transiently undermethylated during replication. In Escherichia coli, undermethylation of the adenine residue in the sequence GATC by the dam gene product helps to recognize the daughter strand where the repair has to be done; otherwise, the error will be permanently fixed (16, 17). Mutants defective in methyl-directed DNA mismatch repair show a strong bias for transition over transversion mismatches (13, 25), and most of the transition mismatches occurring during DNA replication are repaired by this repair pathway.

In E. coli, a very short patch (VSP) repair including the dcm and vsr gene functions along with those of mutS, mutL, and polA has been reported. The T-G mismatches generated in resting cells through spontaneous deamination of 5-methylcytosine to thymine are corrected (14). Nucleotide sequence homologues of E. coli dcm and vsr have been reported in several enteric pathogens, including Shigella sonnei, Salmonella typhimurium, Salmonella enteritidis, Enterobacter cloacae, and Klebsiella pneumoniae (14).

In the course of our studies on DNA mismatch repair in Vibrio cholerae, a gram-negative noninvasive enteric bacterium and the causative agent of the diarrheal disease cholera, several mutator genes involved in methyl-directed mismatch repair, mutL, mutS, and dam, have been identified (5, 6). In the present report, evidence is presented to show that V. cholerae cells lack the dcm methyltransferase gene and the dcm-vsr-mediated VSP repair mechanism. A new gene, mutK, encoding a 19-kDa protein, has been identified which is involved in presumably methyl-independent repair of DNA mismatches. The mutK gene has been cloned, and the complete nucleotide sequence has been determined. The deduced amino acid sequence of mutK showed no significant homology with any of the proteins deposited in the EMBL data bank.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results and discussion References

Bacterial strains and growth conditions. V. cholerae 569B was obtained from the National Institute of Cholera and Enteric Diseases, Calcutta, India. The E. coli strains used in this study were GM31 (thr-1 hisG4 leuB6 rpsL ara-14 supE44 lacY1 tonA31 tsx-78 galK2 galE2 xyl-5 thi-1 mtl-1 dcm-6), XL1-Blue [F::Tn10 proA⁺ B⁺ lacI^q  (lacZ) M15/recA1 endA1 gyrA96 thi hsdR17 (r_K m_K⁺) supE44 relA1], 594 [F lac galK2 galT22 rpsL179 (Str^r) Sup⁰], C600 [F e14 (McrA) thr-1 leuB6 thi-1 lacY1 supE44 rfbD1 fhuA21], CSR603 [thr-1 ara-14 leuB6 (gpt-proA) 62 lacY1 tsx-33 supE44 phr-1 galK2 rac recA1 gyrA98 rpsL31 kdgK51 xyl-5 mtl-1 uvrA6], GW3732 (F thr-1 leuB6 proA2 thi-1 argE3 lacY1 galK2 ara-14 xyl-5 mtl-1 tsx-33 rpsL31 supE44 mutS201::Tn5), GW3734 (F thr-1 leuB6 proA2 his-4 thi-1 argE3 lacY1 galK2 ara-14 xyl-5 mtl-1 tsx-33 rpsL31 supE44 mutL211::Tn5), GW3810 (JM103 dam::Tn9), and SK7755 (F mutU291 rpsL recD1014). E. coli plasmid pDCM2 was obtained from Margaret Lieb (University of Southern California, Los Angeles), and plasmid pNO1575 was provided by K. Ito (Kyoto University, Kyoto, Japan). V. cholerae cells were grown at 37°C in nutrient broth containing 0.1 M NaCl (pH 8.0) and maintained as described by Roy et al. (22). E. coli cells were grown in Luria-Bertani (LB) broth (pH 7.4). Cell viability was assayed as CFU on nutrient broth or LB agar plates.

Preparation of DNA probes and Southern hybridization. DNA was labeled by random priming using the NEBlot kit (New England Biolabs) with [-³²P]dCTP (Amersham, Amersham, United Kingdom). The reaction was carried out at 37°C for 1 h; and the labeled DNA was separated from unincorporated [-³²P]dCTP by passage through a Sephadex G-50 column. For Southern hybridization, about 2 to 3 µg of chromosomal DNA was digested with different restriction enzymes and electrophoresed on a horizontal agarose (1%) slab gel (30 by 13 by 0.5 cm) at 3 V/cm. The gels were stained with ethidium bromide, irradiated with UV light to nick the DNA, denatured, and blotted onto nytran membrane. Hybridization was carried out at 60°C without formamide. The filters were sequentially washed with 3× SSC buffer (1× SSC buffer contains 0.15 M NaCl and 0.015 M Na-citrate) containing 0.5% sodium dodecyl sulfate (SDS) at room temperature and with 2× SSC containing 0.5% SDS at 60°C. The filters were dried and exposed to Kodak XR-5 film by using an intensifying screen.

DNA preparation and spontaneous mutation frequency assay. Chromosomal and plasmid DNAs were isolated by the methods of Brenner et al. (9) and Birnboim and Doly (8), respectively. For measurement of the spontaneous mutation frequency, about 100 cells from an overnight culture were added to 5 ml of LB containing the desired antibiotic and allowed to grow overnight, and the CFU were assayed on LB agar plates. Rifampin-resistant mutants were scored in LB agar plates containing 100 µg of rifampin per ml.

Protein labeling in maxicells. Plasmid-encoded proteins were examined by using maxicell strain CSR603 carrying plasmid pKB370 or pKB130 (23). Transformed cells were exponentially grown (2 × 10⁸ CFU/ml) in minimal medium containing 1% Casamino Acids. The cells were irradiated with UV light (50 J/m²) and incubated at 37°C for 1 h. After 1 h of incubation, 200-µg/ml D-cyloserine was added and incubation was continued for 16 h at 37°C. The cells were harvested, suspended in fresh, sulfur-depleted medium, and incubated for 1 h at 37°C, and the proteins were labeled with [³⁵S]methionine (5 µCi/ml). The labeled cells were harvested, washed, and suspended in electrophoresis sample buffer, and the cell lysate was analyzed by SDS-15% polyacrylamide gel electrophoresis (PAGE), followed by autoradiography of the dried gel (12).

Construction of mutK mutant. A 178-bp HaeIII-HaeIII fragment from the coding region of the V. cholerae mutK gene in pKB130 was cloned into suicide vector pGP704 (19), and recombinant vector pKS178 was maintained in E. coli SM10. Recombinant plasmid pKS178 was conjugally transferred from E. coli SM10 cells into streptomycin-resistant V. cholerae cells. Conjugants were isolated on streptomycin (150 µg/ml)- and ampicillin (50 µg/ml)-containing nutrient agar plates and screened for cells in which the plasmid was integrated into the chromosome. This was further confirmed by Southern blot hybridization of BamHI- or EcoRI-digested genomic DNA by using the 0.81-kb KpnI-BamHI fragment of pKB370 containing the mutK gene of V. cholerae.

DNA sequencing. The DNA sequence was determined by the dideoxynucleotide chain termination method (24). The complete nucleotide sequences of mutK and the genes immediately up- and downstream have been deposited in the EMBL data bank (accession no. Y11983 and Y11908).

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and methods Results and discussion References

While searching for VSP repair genes and their function in V. cholerae, an E. coli vsr gene probe, a 2-kb NcoI-AccI fragment of the 7-kb DNA of E. coli genomic DNA containing the C-terminal end of the vsr gene along with some other sequences, was used. The E. coli probe hybridized with V. cholerae DNA in the 6.5-, 9.3-, and 12-kb regions following digestion with EcoRI, PstI, and BamHI, respectively. A minibank was constructed from the 6.5-kb region of EcoRI-digested chromosomal DNA in vector pNO1575, and the bank was maintained in E. coli XL1-Blue. From about 650 recombinant colonies, plasmids were isolated, transferred to a nylon membrane, and hybridized with the E. coli probe. Three recombinant plasmids hybridized with the probe, and all of them contained a 6.7-kb DNA fragment. Digestion of the 6.7-kb DNA fragment with BamHI produced three fragments with sizes of 3.7, 1.4, and 1.6 kb; only the 3.7-kb fragment hybridized with the E. coli probe.

A physical map of the 3.7-kb DNA fragment was constructed by using several enzymes (Fig. 1). The 1.3-kb NcoI-NcoI fragment of the 3.7-kb DNA that hybridized with the E. coli probe was cloned at the NcoI site of pACYC184 and pUC19, generating recombinant plasmids pKB130 and pKB131, respectively (Fig. 1). The 3.7-kb DNA fragment was also cloned either into EcoRI-BamHI-digested vector pNO1575, generating recombinant plasmid pKB222, or into the vector pUC19, resulting in the plasmid pKB370 (Fig. 1).

View larger version (25K): [in this window] [in a new window]   FIG. 1.   Physical map of the 3.7-kb V. cholerae DNA carrying the mutK gene (solid bar) and construction of recombinant plasmids pKB222, pKB370, pKB130, pKB131, and pKS80. The horizontal arrows show the direction and extent of sequencing. The thick horizontal arrows denote the ORFs present in the 3.7-kb DNA. The thin line represents the vector sequences, and the thick line represents the chromosomal insert. A, AseI; B, BamHI; E, EcoRI; H, HindIII; K, KpnI; M, MluI; N, NcoI; X, XmnI.

To examine whether the 3.7- or 1.3-kb V. cholerae DNA can functionally complement E. coli vsr mutants, strain GM31 (dcm-6) was transformed with plasmid pKB370 or pKB130 and VSP repair was examined by the bacteriophage -based assay (14, 26). Neither the 3.7-kb nor the 1.3-kb DNA could complement the phenotypic traits associated with the E. coli vsr mutant. The hybridization of V. cholerae genomic DNA with the E. coli vsr probe was due not to the vsr gene but to a small gene whose function is not known that is present in the vsr probe. Thus, V. cholerae does not have an E. coli vsr homologue. V. cholerae chromosomal DNA and plasmid DNAs isolated from V. cholerae were sensitive to digestion by the enzyme EcoRII, suggesting that there is no dcm gene in V. cholerae. This was confirmed by the lack of hybridization of an E. coli dcm gene probe to the V. cholerae chromosomal DNA.

Proteins encoded by the 3.7- and 1.3-kb DNA fragments. The protein(s) encoded by the 3.7- and 1.3-kb V. cholerae DNA fragments was examined in maxicells (23). Cells of E. coli CSR603 carrying pKB370 were irradiated with UV light and labeled with [³⁵S]methionine for 1 h, and the soluble-protein extract was analyzed by SDS-PAGE. The 3.7-kb DNA fragment encodes two proteins with molecular masses of 14 and 19 kDa, and the 14-kDa protein is produced in large amounts (Fig. 2). The 1.3-kb DNA fragment encodes only the 19-kDa protein.

View larger version (42K): [in this window] [in a new window]   FIG. 2.   Identification of proteins encoded by the 3.7-kb V. cholerae DNA fragment. UV-irradiated E. coli CSR603 carrying plasmid pUC19 (lane a) or pKB370 (lane b) was labeled with [35S]methionine, and plasmid-encoded proteins were examined by SDS-PAGE. The same amount of radioactivity was loaded in each lane. The values are molecular masses of proteins in kilodaltons.

The 19-kDa protein can partially suppress the repair-defective phenotype of E. coli mutS, mutL, and mutU mutants. Since the 19-kDa protein was initially suspected to be involved in DNA mismatch repair, recombinant plasmids pKB370, pKB130, and pKB131 were transformed into E. coli mutS (GW3732), mutL (GW3734), and mutU (SK7755) mutants respectively, and spontaneously occurring rifampin-resistant mutants were scored. Surprisingly, although the 19-kDa protein is not a VSR analogue, it could partially suppress the phenotypic traits associated with E. coli mutS, mutL, and mutU mutants. An about 10-fold reduction in the spontaneous mutation frequency in the E. coli mutS and mutL mutants and a 6-fold reduction in the mutU mutant was observed in the presence of either the 3.7- or 1.3-kb V. cholerae genomic DNA (Table 1). The reduction in spontaneous mutation was directly proportional to the amount of the 19-kDa protein in the cell. When the 3.7- or 1.3-kb DNA was cloned into vector pACYC184 instead of pUC19, the reduction in the spontaneous mutation frequency was less but was highly reproducible. The effect is pronounced in mutS mutants. When the 1.3-kb NcoI-NcoI fragment was deleted from pKB370, it failed to complement E. coli mutS, mutL, and mutU mutants. Thus, the 19-kDa protein, having no sequence homology with the mutS, mutL, or mutU gene, has an antimutagenic property and can partially suppress the phenotypic traits associated with E. coli mutS, mutL, and mutU mutants.

Nucleotide and deduced amino acid sequences of the 3.7-kb DNA. The nucleotide sequence of the 3.7-kb V. cholerae DNA has been determined. The deduced amino acid sequence showed three open reading frames (ORFs) of 70 (ORF1; sequence not shown; EMBL accession no. Y11908), 140 (ORF2), and 163 (ORF3) (Fig. 3) amino acid residues. A protein data bank search using BLASTX (2) revealed that ORF1 has 80% homology at the amino acid level with a highly conserved major cold shock protein (CspA) found in several bacteria. This protein has extensive homology with human DNA binding proteins, which suggests that it must have some function and can no longer be regarded as hypothetical. In V. cholerae, this protein is located in the 5' region of ORF3. In the absence of both dcm and vsr sequences in V. cholerae, the E. coli vsr gene probe used hybridized with genomic DNA because of the presence of ORF2. It is likely that the mutK gene is located in the region into which the dcm and vsr genes were inserted in E. coli.

View larger version (70K): [in this window] [in a new window]   FIG. 3.   Nucleotide sequence of the 1.55-kb MboII-BamHI fragment of the 3.7-kb V. cholerae DNA and deduced amino acid sequences of ORF2, ORF3, and ORF4. Sequencing was done by the dideoxy-chain termination method (33). The nucleotides and amino acids are numbered on the right. *, stop codon; S.D, putative ribosome binding site.